You are here
Ronan0 - Wed, 2014/03/12 - 15:51
I need to increase the size of the volume on a Turnkey EC2 microinstance of vTiger (production instance).
Can I confirm the procedure here:
1. Stop the instance.
2. Take a snapshot of the volume.
3. Recreate a new volume from the snapshot of increased size.
4. Detach old volume and reattach new volume (heeding mount points.
5. Restart the instance.
6. Update linux to recognise the increased volume size with command: resize2fs /dev/xvdal
Any pointers? Will Tklbam be affected? What should I look out for? Do I need to deal with LVM in Webmin?
Please note that I am uncomfortable using ephemeral storage in case I need to stop the instance.
TIA.
Forum:
TBH I'm not sure...
Regardless, TKLBAM shouldn't be affected as that operates on the mounted filesystem hierarchy (irregardless of the underlying volumes).
Any progress on this? Trying to understand EBS volumes myself.
If the Bronze Level is subscribed to what specifically is just the method for dealing with "EBS Backing"? Is it standard? It seems S2 (or whatever the code is) is standard. Can an EBS volume be spedified at setup or do data and files need to be migrated to it.
It seems that EBS volumes can be created quite readily so I would think TKLBAM would be a method for migrating to a larger EBS volume. I need to spend some time on how still, however.
So my comments are more of questions on a different subject. Perhaps should have been a diff thread but this seemed to be hanging a bit and recent when I searched for info.
Yes EBS backing is standard on all plans
Except the "Pay-per-use" (which is S3 backed). You actually need a "Pay-per-use" plan to launch an S3 backed server, none of the other plans support it...
Yes TKLBAM can be used to migrate from one server to another, regardless of where the server is running or what type of storage it has. FWIW EBS backed servers also support 'snapshots', although they are not recommended to be used as your recovery backup (although can be handy and quick; it is still recommended that you use TKLBAM for your backups).
To update my answer to the original post, I would think that the process outlined should work fine, although my inclination would be to just use TKLBAM. It would be a good opportunity to test the backup at the same time I would think...
LVM is not supported nor used on AWS instance AFAIK (although TBH I've never bothered checking...).
If EBS is standard what are the small SSD drives specified
On this Page there are a number of configurations noted.
https://hub.turnkeylinux.org/amazon/sizes/help/
Small have 160 G Drives
Medium have 4 G SSD
Meanwhile I have read that EBS can be slower than S3 etc. Trying to understand the Amazon system is all. I've been running my own VmWare on esxi and the only correlation I know is Hard Drive size. Seems there is a diff between Old and New drives and that buying a Reserved instance is essentially buying some Hard Drives (or Solid State Drives) for Amazon... Still want to know what real differences there are in Hub instances and variables that can and cannot be enabled.
(more and more off subject I expect)
They are what's called 'instance storage'
TBH I'm not sure whether it's basically like S3 storage or not, but it's basically storage space that comes with your server at no additional charge. By default it's mounted to /mnt/tmp and you can use mount-bind to mount it to any part of your filesystem you please. From what I understand it is generally slower than EBS (although perhaps not the SSD backed ones?) but comes with your instance free (whereas you pay for the EBS - although not a lot in my experience).
I don't think EBS is slower than S3 but I've never benchmarked it... From my experience they perform fairly similarly, although I haven't used an S3 backed instance for a few years now...
Reserved Instances are basically just pre-paying for your server(s). I think that I have probably already pretty much covered it, but to be totally clear (and in case someone else reads this thread and not your other) how it works is; you pay an upfront fee and in exchange, you get a much cheaper hourly rate. Once you know how many, what size and which region you want your servers to use then Reserved Instances can save you plenty of money. But if you don't really know what you want/need then you might end up pre-paying for something that you don't want/need and they aren't transferable or refundable...
Add new comment